The lesson we learned yesterday during the demo in Saratov is the following: never ever give demos to a crowd of journalists all together in the vehicle! Why? Well, I’ll take a step back and tell you the full story.
Saratov. VisLab is preparing the demo. Everything works flawlessly. We decide to prepare a waypoint following demo: nothing special, but it is generally well received by journalists and spectators. It works: the vehicle moves in the area and follows a prerecorded set of GPS positions. Then the journalists come and ask for the strangest things (they always want some better image than their colleagues…). The vehicle is behaving so nicely that we decide to give a demo with journalists in front (driving seat and passenger seat) and their cameramen in the back seats. One of us is squeezed in one of the back seats to launch the demo and to finish it (with a laptop on his legs and quite limited movements). The demo starts. Everything goes well. When the demo is about to end, something strange happens and the vehicle does not stop at the last waypoint and continues moving: the speed constantly decreases but the vehicle keeps on moving until it touches another vehicle. Our staff was speechless, that was an unbelievable move of the vehicle, very strange and unpredicted. The media gets off and we don’t understand what prevented the vehicle from stopping at the last waypoint. Then we check the vehicle and it seems that one journalist (hopefully accidentally), maybe with his/her elbow or with the camera, kept pushing on the x-by-wire button, switching it off and on again a few times, thus the actuation was inhibited and the vehicle could not either steer or brake.
We are now going to analyze the full log to understand when this happened and what exactly are the consequences of this switching on and off, but whether this will be confirmed or not, we have learned that we will NEVER EVER give demos to many passengers at a time again. We will have at least two VisLab engineers on board: one to check the vehicle’s behavior, and one to check the passenger’s behavior!
Just for our readers which are not familiar with the vehicle setup, here’s a picture of the armrest in the back seat: it was designed so that all subsystems were easily reachable by our staff during the test, but it turns out that they are definitely too easily reachable also by other people during the demos.
This time we are convinced the problem is not due to a design issue but to a naive behavior of our crew: the vehicle is a test vehicle and these switches must be handy and placed in a convenient position.
Well, our lesson learned now is that we will now have to be careful about not only the vehicle but the passengers as well!